US6728319B1ExpiredUtility

Decoding method and apparatus

Assignee: ERICSSON TELEFON AB L MPriority: Dec 22, 1997Filed: Dec 17, 1998Granted: Apr 27, 2004
Est. expiryDec 22, 2017(expired)· nominal 20-yr term from priority
H03M 7/3082
46
PatentIndex Score
12
Cited by
14
References
22
Claims

Abstract

A decoder in a communication system using channel optimized vector quantization and including an encoder stores active encoder centroids c i corresponding to active indices i, an active index being defined as an index that has a predetermined source probability p i >0 of being selected by the encoder for transmission to the decoder, source probabilities p i and the bit error rate epsilon to dynamically estimate a decoded vector using a sub-optimal algorithm in which only the most significant contributions are considered.

Claims

exact text as granted — not AI-modified
What is claimed is:  
     
       1. A decoding method in a communication system using channel optimized vector quantization and including an encoder and a decoder, characterized by performing the following steps at said decoder: 
       storing active encoder centroids c i  corresponding to active indices i, an active index being defined as an index that has a predetermined source probability p i >0 of being selected by said encoder for transmission to said decoder over a transmission channel;  
       storing said source probabilities p i ;  
       storing a transmission channel quality measure;  
       dynamically calculating, from said stored transmission channel quality measure, a set of conditional channel transition probabilities p j|i  that an index j has been received by said decoder given that an index i has been sent by said encoder;  
       calculating an estimate ĉ j  of a decoded vector in accordance with the formula.            c   ^     j     =         ∑       active   indices     i                                 c   i     ·     p     j   |   i       ·     p   i             ∑       active   indices     i                                 p     j   |   i       ·     p   i                           
       where the summations are restricted to only the most significant terms, and the terms are summed in descending conditional channel transition probability order; and 
       stopping summation at the latest when a predetermined number L, which is less than the number of active indices, of terms have been summed.  
     
     
       2. The method of  claim 1 , characterized by calculating said estimate ĉ j  of a decoded vector in accordance with the formula:            c   ^     j     ≈         ∑         active   indices       i                 for                 which           p   i     >   δ                                   c   i     ·     p     j   |   i       ·     p   i             ∑         active   indices       i                 for                 which           p   i     >   δ                                   p     j   |   i       ·     p   i                           
       where δ is a predetermined small positive constant. 
     
     
       3. The method of  claim 1 , characterized by 
       accumulating the conditional channel transition probabilities of summed terms; and  
       stopping summation when a predetermined probability threshold has been exceeded by the accumulated conditional channel transition probabilities before L terms have been summed.  
     
     
       4. The method of any of the preceding claims  1 ,  2 , or  3 , characterized by said transmission channel quality measure being the bit error rate ε of said transmission channel. 
     
     
       5. The method of  claim 4 , characterized by calculating said conditional channel transition probabilities p j|i  in accordance with the formula 
       
         
             p   j|i =ε d     H     (j,i) (1−ε) N−d     H     (j,i)    
         
       
       where 
       N denotes the number of bit positions in an index, and  
       d H (j,i) denotes the Hamming distance between j and i.  
     
     
       6. A decoding method in a communication system using channel optimized vector quantization and including an encoder and a decoder, characterized by performing the following steps at said decoder: 
       storing active encoder centroids c i  corresponding to active indices i, an active index being defined as an index that has a predetermined source probability p i >0 of being selected by said encoder for transmission to said decoder over a transmission channel;  
       storing said source probabilities p i ;  
       storing a transmission channel quality measure;  
       dynamically calculating, from said stored transmission channel quality measure a set of conditional channel transition probability density function values f R|I (r|i) representing the probability density that a soft decoded index r has been received by said decoder given that an index i has been sent by said encoder;  
       calculating an estimate ĉ j  of a decoded vector in accordance with the formula:            c   ^       r   _       =         ∑       active   indices     i                                   c   i     ·     f     R   |   I                           (       r   _     |   i     )     ·     p   i               ∑       active   indices     i                                 f     R   |   I                         (       r   _     |   i     )     ·     p   i                             
       where the summations are restricted to only the most significant terms, and the terms are summed in descending conditional channel transition probability density function value order; and 
       stopping summation when a predetermined number L, which is less than the number of active indices, of terms have been summed.  
     
     
       7. The method of  claim 6 , characterized by calculating said estimate ĉ r  of said decoded vector in accordance with the formula:            c   ^       r   _       ≈         ∑         active   indices       i                 for                 which           p   i     >   δ                                     c   i     ·     f     R   |   I                           (       r   _     |   i     )     ·     p   i               ∑         active   indices       i                 for                 which           p   i     >   δ                                   f     R   |   I                         (       r   _     |     \      i       )     ·     p   i                             
       where δ is a predetermined small positive constant. 
     
     
       8. The method of  claim 6 , characterized by 
       accumulating the conditional channel transition probability density function values of summed terms; and  
       stopping summation when a predetermined probability threshold has been exceeded by the accumulated conditional channel transition probability density function values before L terms have been summed.  
     
     
       9. The method of any of the preceding claims  claim 6 ,  7 , or  8 , characterized by said transmission channel quality measure being a noise variance σ 2  of said transmission channel. 
     
     
       10. The method of  claim 9 , characterized by calculating said conditional channel transition probability density function values f R|I (r|i) in accordance with the formula            f     R   |   I                       (       r   _     |   i     )       =       1       2                 π                 σ                                -     1     2                   σ   2                                    r   _     -       b   _                     (   i   )              2                           
       where 
       b(i) represents an index i as a vector in which each binary digit in i corresponds to either +1 or −1; and  
       ∥·∥ 2  denotes squared euclidean distance.  
     
     
       11. A decoding apparatus in a communication system using channel optimized vector quantization and including an encoder and a decoder, characterized by: 
       means for storing active encoder centroids c i  corresponding to active indices i, an active index being defined as an index that has a predetermined source probability p i >0 of being selected by said encoder for transmission to said decoder over a transmission channel;  
       means for storing said source probabilities p i ;  
       means for storing a transmission channel quality measure;  
       means for dynamically calculating, from said stored transmission channel quality measure, a set of conditional channel transition probabilities p j|i  that an index j has been received by said decoder given that an index i has been sent by said encoder;  
       means for calculating an estimate ĉ j  of a decoded vector in accordance with the formula:            c   ^     j     =         ∑       active   indices     i                                 c   i     ·     p     j   |   i       ·     p   i             ∑       active   indices     i                                 p     j   |   i       ·     p   i                           
       where the summations are restricted to only the most significant terms, and the terms are summed in descending conditional channel transition probability order: and 
       means for stopping summation at the latest when a predetermined number L, which is less than the number of active indices, of terms have been summed.  
     
     
       12. The apparatus of  claim 11 , characterized by means for calculating said estimate ĉ j  of a decoded vector in accordance with the formula:            c   ^     j     ≈         ∑         active   indices       i                 for                 which           p   i     >   δ                                   c   i     ·     p     j   |   i       ·     p   i             ∑         active   indices       i                 for                 which           p   i     >   δ                                   p     j   |   i       ·     p   i                           
       where δ is a predetermined small positive constant. 
     
     
       13. The apparatus of  claim 11 , characterized by 
       means for accumulating the conditional channel transition probabilities of summed terms; and  
       means for stopping summation when a predetermined probability threshold has been exceeded by the accumulated conditional channel transition probabilities before L terms have been summed.  
     
     
       14. The apparatus of any of the preceding claims  11 ,  12 , or  13 , characterized by means for calculating said conditional channel transition probabilities p j|i  in accordance with the formula 
       
         
             p   j|i =ε d     H     (j,i) (1−ε) N−d     H     (j,i)    
         
       
       where 
       N denotes the number of bit positions in an index,  
       d H (j,i) denotes the Hamming distance between j and, and  
       ε denotes the bit error rate of said transmission channel.  
     
     
       15. A decoding apparatus in a communication system using channel optimized vector quantization and including an encoder and a decoder, characterized by: 
       means for storing active encoder centroids c i  corresponding to active indices i, an active index being defined as an index that has a predetermined source probability p i >0 of being selected by said encoder for transmission to said decoder over a transmission channel;  
       means for storing said source probabilities p i ;  
       means for storing a transmission channel quality measure;  
       means for dynamically calculating, from said stored transmission channel quality measure, a set of conditional channel transition probability density function values f R|I (r|i) representing the probability density that a soft decoded index r has been received by said decoder given that an index i has been sent by said encoder;  
       means for calculating an estimate ĉ j  of a decoded vector in accordance with the formula:            c   ^       r   _       =         ∑       active   indices     i                                   c   i     ·     f     R   |   I                           (       r   _     |   i     )     ·     p   i               ∑       active   indices     i                                 f     R   |   I                         (       r   _     |   i     )     ·     p   i                             
       where the summations are restricted to only the most significant terms, and the terms are summed in descending conditional channel transition probability density function value order; and 
       means for stopping summation when a predetermined number L, which is less than the number of active indices, of terms have been summed.  
     
     
       16. The apparatus of  claim 15 , characterized by means for calculating said estimate ĉ r  of said decoded vector in accordance with the formula:            c   ^       r   _       ≈         ∑         active   indices       i                 for                 which           p   i     >   δ                                     c   i     ·     f     R   |   I                           (       r   _     |   i     )     ·     p   i               ∑         active   indices       i                 for                 which           p   i     >   δ                                   f     R   |   I                         (       r   _     |   i     )     ·     p   i                             
       where δ is a predetermined small positive constant. 
     
     
       17. The apparatus of  claim 15 , characterized by 
       means for accumulating the conditional channel transition probability density function values of summed terms; and  
       means for stopping summation when a predetermined probability threshold has been exceeded by the accumulated conditional channel transition probability density function values before L terms have been summed.  
     
     
       18. The apparatus of any of the preceding claims  15 ,  16 , or  17 , characterized by means for calculating said conditional channel transition probability density function values f R|I (r|i) in accordance with the formula            f     R   |   I                       (       r   _     |   i     )       =       1       2                 π                 σ                                -     1     2                   σ   2                                    r   _     -       b   _                     (   i   )              2                           
       where 
       b(i) represents an index i as a vector in which each binary digit in i corresponds to either +1 or −1,  
       ∥·∥ 2  denotes squared euclidean distance, and  
       σ 2  denotes a noise variance of said transmission channel.  
     
     
       19. A mobile station in a communication system using channel optimized vector quantization, characterized by a decoder including: 
       means for storing active encoder centroids c i  corresponding to active indices i, an active index being defined as an index that has a predetermined source probability p i >0 of being selected by an encoder for transmission to said decoder over a transmission channel;  
       means for storing said source probabilities p i ;  
       means for storing a transmission channel quality measure;  
       means for dynamically calculating, from said stored transmission channel quality measure, a set of conditional channel transition probabilities p j|i  that an index j has been received by said decoder given that an index i has been sent by said encoder,  
       means for calculating an estimate ĉ i  of a decoded vector in accordance with the formula:            c   ^     j     =         ∑       active   indices     i                                 c   i     ·     p     j   |   i       ·     p   i             ∑       active   indices     i                                 p     j   |   i       ·     p   i                           
       where the summations are restricted to only the most significant terms, and the terms are summed in descending conditional channel transition probability order; and 
       means for stopping summation when a predetermined number L, which is less than the number of active indices, of terms have been summed.  
     
     
       20. A mobile station in a communication system using channel optimized vector quantization, characterized by a decoder including: 
       means for storing active encoder centroids c i  corresponding to active indices i, an active index being defined as an index that has a predetermined source probability p i >0 of being selected by an encoder for transmission to said decoder over a transmission channel;  
       means for storing said source probabilities p i ;  
       means for storing a transmission channel quality measure;  
       means for dynamically calculating, from said stored transmission channel quality measure, a set of conditional channel transition probability density function values f R|I (r|i) representing the probability density that a soft decoded index r has been received by said decoder given that an index i has been sent by said encoder;  
       means for calculating an estimate ĉ j  of a decoded vector in accordance with the formula:            c   ^       r   _       =         ∑       active   indices     i                                   c   i     ·     f     R   |   I                           (       r   _     |   i     )     ·     p   i               ∑       active   indices     i                                 f     R   |   I                         (       r   _     |   i     )     ·     p   i                             
       where the summations are restricted to only the most significant terms, and the terms are summed in descending conditional channel transition probability density function value order; and 
       means for stopping summation when a predetermined number L, which is less than the number of active indices, of terms have been summed.  
     
     
       21. A base station in a communication system using channel optimized vector quantization, characterized by a decoder including: 
       means for storing active encoder centroids c i  corresponding to active indices i, an active index being defined as an index that has a predetermined source probability p i >0 of being selected by an encoder for transmission to said decoder over a transmission channel;  
       means for storing said source probabilities p i ;  
       means for storing a transmission channel quality measure;  
       means for dynamically calculating, from said stored transmission channel quality measure, a set of conditional channel transition probabilities p j|i  that an index j has been received by said decoder given that an index i has been sent by said encoder;  
       means for calculating an estimate ĉ j  of a decoded vector in accordance with the formula:            c   ^     j     =         ∑       active   indices     i                                 c   i     ·     p     j   |   i       ·     p   i             ∑       active   indices     i                                 p     j   |   i       ·     p   i                           
       where the summations are restricted to only the most significant terms, and the terms are summed in descending conditional channel transition probability order; and 
       means for stopping summation when a predetermined number L, which is less than the number of active indices, of terms have been summed.  
     
     
       22. A base station in a communication system using channel optimized vector quantization, characterized by a decoder including: 
       means for storing active encoder centroids c i  corresponding to active indices i, an active index being defined as an index that has a predetermined source probability p i >0 of being selected by an encoder for transmission to said decoder over a transmission channel;  
       means for storing said source probabilities p i ;  
       means for storing a transmission channel quality measures  
       means for dynamically calculating, from said stored transmission channel quality measure, a set of conditional channel transition probability density function values f R|I (r|i) representing the probability density that a soft decoded index r has been received by said decoder given that an index i has been sent by said encoder:  
       means for calculating an estimate ĉ j  of a decoded vector in accordance with the formula:            c   ^       r   _       =         ∑       active   indices     i                                   c   i     ·     f     R   |   I                           (       r   _     |   i     )     ·     p   i               ∑       active   indices     i                                 f     R   |   I                         (       r   _     |   i     )     ·     p   i                             
       where the summations are restricted to only the most significant terms, and the terms are summed in descending conditional channel transition probability density function value order; and 
       means for stopping summation when a predetermined number L, which is less than the number of active indices, of terms have been summed.

Join the waitlist — get patent alerts

Track US6728319B1 — get alerts on status changes and closely related new filings.

We store only your email — no account needed. See our privacy policy.